List your company

Task Success on ALFWorld (test)

94.23Success RateMetaSkill-Evolve
90.22691.265592.30593.3445Jul 6, 2026

Evaluation results

MethodLinks
MetaSkill-EvolveBackbone=Gemma-4 31B2026.0794.23
No-SkillBackbone=Gemma-4 31B2026.0792.31
Single-LevelBackbone=Gemma-4 31B2026.0792.31
Static skillBackbone=Gemma-4 31B2026.0790.38